Hangchen Y
Senior Staff Research Scientist at Meta
Menlo Park, California, United States
Summary
🤩
Rockstar🎓
Top SchoolHangchen Y is a Senior Staff Research Scientist at Meta with 11 years of experience building high-performance ML and systems software. He holds a PhD in Computer Science from UT Austin and combines deep research rigor with hands-on engineering across compilers, performance, and recommendation systems. At Meta he progressed from Staff to Senior Staff, driving production-ready research and scalable infrastructure for large models. An active open-source contributor, he has improved the PyTorch Glow compiler—enhancing error reporting, robustness, multi-card Recsys execution, and caching for hardware accelerators. Colleagues rely on him for pragmatic solutions that bridge research experiments and production constraints, and he’s known for quietly improving developer velocity through better tooling and diagnostics. Based in Menlo Park, he blends academic depth with practical impact on real-world ML workloads.
11 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at The University of Texas at AustinBachelor of Science - BS Electrical Engineering, Bachelor of Science - BS Electrical Engineering at Shanghai Jiao Tong University
Stackoverflow
335reputation
19kreached
12answers
5questions
Github Skills (16)
error-handling10
compiler-design10
c-language10
cprogramming-language10
back-end-development10
llvm10
machine-learning9
recommendation-system9
multithreading8
algorithm6
deep-learning6
tensorflow6
loops6
smtp6
python6
Programming languages (6)
Github contributions (5)
pytorch/glow
Apr 2021 - Apr 2022
Compiler for Neural Network hardware accelerators
Role in this project:
Back-end Developer
Back-end DeveloperContributions:10 commits, 3 PRs in 1 year
Contributions summary:Hangchen contributed to the Glow compiler by improving error messages, introducing macros for error reporting, and enabling Recsys models. They focused on enhancing the caching mechanism and making the replayer more robust by handling potential errors during resource creation. Furthermore, the user enabled the execution of Recsys models on multiple cards and also incorporated features for host saturation. They also replaced logging information with VLOGs and fixed CI test breaks.
hardware-acceleratorscompilerneural-networkacceleratorshardware
yuhc/ahk-script
Jan 2015 - May 2020
Contributions:3 commits, 2 pushes in 5 years 5 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.